Expand description
AxonML Vision — Computer Vision Crate
Top-level crate module for the axonml-vision computer vision library.
Re-exports image transforms (Resize, CenterCrop, ColorJitter, Grayscale,
ImageNormalize, Pad, RandomHorizontalFlip, RandomRotation, RandomVerticalFlip,
ToTensorImage), datasets (MNIST, CIFAR10, CIFAR100, FashionMNIST, and their
synthetic variants), classification models (LeNet, MLP, SimpleCNN), biometric
identity types (AegisIdentity, Mnemosyne, Ariadne, Echo, Argus, Themis), and
pretrained model hub utilities. Provides a prelude module for common imports.
§File
crates/axonml-vision/src/lib.rs
§Author
Andrew Jewell Sr. — AutomataNexus LLC ORCID: 0009-0005-2158-7060
§Updated
April 16, 2026 11:15 PM EST
§Disclaimer
Use at own risk. This software is provided “as is”, without warranty of any kind, express or implied. The author and AutomataNexus shall not be held liable for any damages arising from the use of this software.
Re-exports§
pub use transforms::CenterCrop;pub use transforms::ColorJitter;pub use transforms::Grayscale;pub use transforms::ImageNormalize;pub use transforms::Pad;pub use transforms::RandomHorizontalFlip;pub use transforms::RandomRotation;pub use transforms::RandomVerticalFlip;pub use transforms::Resize;pub use transforms::ToTensorImage;pub use datasets::CIFAR10;pub use datasets::CIFAR100;pub use datasets::FashionMNIST;pub use datasets::MNIST;pub use datasets::SyntheticCIFAR;pub use datasets::SyntheticMNIST;pub use models::LeNet;pub use models::MLP;pub use models::SimpleCNN;pub use models::biometric::AegisIdentity;pub use models::biometric::AngularMarginLoss;pub use models::biometric::ArgusIris;pub use models::biometric::ArgusLoss;pub use models::biometric::AriadneFingerprint;pub use models::biometric::BiometricConfig;pub use models::biometric::BiometricEvidence;pub use models::biometric::BiometricModality;pub use models::biometric::CenterLoss;pub use models::biometric::ContrastiveLoss;pub use models::biometric::CrystallizationLoss;pub use models::biometric::DiversityRegularization;pub use models::biometric::EchoLoss;pub use models::biometric::EchoSpeaker;pub use models::biometric::EnrollmentResult;pub use models::biometric::IdentificationResult;pub use models::biometric::IdentityBank;pub use models::biometric::LivenessLoss;pub use models::biometric::MnemosyneIdentity;pub use models::biometric::ThemisFusion;pub use models::biometric::ThemisLoss;pub use models::biometric::VerificationResult;pub use hub::HubError;pub use hub::HubResult;pub use hub::PretrainedModel;pub use hub::StateDict;pub use hub::cache_dir;pub use hub::download_weights;pub use hub::is_cached;pub use hub::list_models;pub use hub::load_state_dict;pub use hub::model_info;pub use hub::model_registry;
Modules§
- camera
- Camera Capture and Inference Pipeline — Core Types
- datasets
- Vision Datasets — Dataset Module Re-exports
- edge
- Edge Deployment Pipeline — Model Profiling and Edge Registry
- hub
- Model Hub — Pretrained Weights Registry and Download
- image_
io - Image I/O — Load JPEG/PNG/BMP Images into Tensors
- losses
- Detection Losses — Focal Loss, GIoU Loss, Uncertainty NLL Loss
- models
- Vision Models — Model Module Re-exports and Smoke Tests
- ops
- Vision Operations — Detection Primitives, NMS, Interpolation, RoI Align
- prelude
- Common imports for computer vision tasks.
- training
- Detection Training Module — FCOS Assignment and Per-Step Loss Runners
- transforms
- Image Transforms — Vision-Specific Data Augmentation Pipeline